Merge pull request #121 from sysuyy/add_multi_node_script
[feat]: use accelerate on multi-node Former-commit-id: 67b01dc1d934d3d86d7ecca00c5dd1a11fd4d22d
This commit is contained in:
@@ -0,0 +1,2 @@
|
||||
server1 slots=8 # your server name and GPU in total
|
||||
server2 slots=8
|
||||
@@ -0,0 +1,18 @@
|
||||
compute_environment: LOCAL_MACHINE
|
||||
distributed_type: DEEPSPEED
|
||||
deepspeed_config:
|
||||
deepspeed_config_file: scripts/accelerate_configs/zero3_offload.json
|
||||
deepspeed_hostfile: # FullPath here
|
||||
fsdp_config: {}
|
||||
machine_rank: 0
|
||||
main_process_ip: # main process ip
|
||||
main_process_port: 29501
|
||||
main_training_function: main
|
||||
num_machines: 2
|
||||
num_processes: 16
|
||||
rdzv_backend: static
|
||||
same_network: true
|
||||
tpu_env: []
|
||||
tpu_use_cluster: false
|
||||
tpu_use_sudo: false
|
||||
use_cpu: false
|
||||
Reference in New Issue
Block a user